Filetype PDF | Posted on 09 Jan 2023 | 2 years ago
Authentication
digital resources
158x Filetype PDF File size 0.39 MB Source: apjcn.nhri.org.tw
756 Asia Pac J Clin Nutr 2018;27(4):756-762
Original Article
Development of a screening tool to detect nutrition risk
in patients with inflammatory bowel disease
1 1,2
Natasha Haskey RD, MSc , Juan Nicolás Peña-Sánchez MD, MPH, PhD ,
3 1
Jennifer L Jones MD, MSc , Sharyle A Fowler MD
1
Multidisciplinary Inflammatory Bowel Disease Clinic, Division of Gastroenterology, Department of
Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
2
Department of Community Health and Epidemiology, University of Saskatchewan, Saskatoon, Saskatchewan,
Canada
3
Division of Digestive Care & Endoscopy, Department of Medicine, Dalhousie University, Halifax, Nova
Scotia, Canada
Background and Objectives: Malnutrition is a known complication of Inflammatory Bowel Disease (IBD). We
assessed a known screening tool, as well as developed and validated a novel screening tool, to detect nutrition
risk in outpatients with IBD. Methods and Study Design: The Saskatchewan IBD–Nutrition Risk (SaskIBD-NR
Tool) was developed and administered alongside the Malnutrition Universal Screening Tool (MUST). Nutrition
risk was confirmed by the IBD dietitian (RD) and gastroenterologist (GI). Agreement between screening tools
and RD/GI assessment was computed using Cohen’s kappa. Results: Of the 110 patients screened, 75 (68.2%)
2
patients had Crohn’s Disease and 35 (31.8%) ulcerative colitis. Mean BMI was 26.4 kg/m (SD=5.8). RD/GI as-
sessment identified 23 patients (20.9%) at nutrition risk. The SaskIBD-NR tool classified 21 (19.1%) at some nu-
trition risk, while MUST classified 17 (15.5%). The SaskIBD-NR tool had significant agreement with the RD/GI
assessment (k 0.83, p<0.001), while MUST showed a lack of agreement (k 0.15, p=0.12). The SaskIBD-NR had
better sensitivity (82.6% vs 26.1%), specificity (97.7% vs 87.4%), positive predictive value (90.5% vs 35.3%),
and negative predictive value (95.5% vs 81.7%) than the MUST. Conclusion: The SaskIBD-NR, which assesses
GI symptoms, food restriction, and weight loss, adequately detects nutrition risk in IBD patients. Broader valida-
tion is required.
Key Words: nutrition screening tool, nutrition risk, inflammatory bowel disease, Crohn’s disease, ulcerative colitis
INTRODUCTION and undertreatment of both malnutrition and nutrient de-
1,14
Malnutrition and weight loss are well recognized compli- ficiencies. The Subjective Global Assessment (SGA) is
1
cations of inflammatory bowel disease (IBD). One of the widely considered the gold standard to diagnose patients
15
most under recognized mechanisms of malnutrition is who are moderately or severely malnourished. However,
reduced food intake and specific avoidance of foods patients who are at risk of malnutrition should be identi-
among IBD patients. Up to 90% of Crohn’s disease (CD) fied early when interventions can be applied, rather than
patients and 71% of ulcerative colitis (UC) patients in once they are already malnourished. SGA does not assess
2
remission use elimination diets to control symptoms. whether patients are avoiding food items or food groups.
Protein-energy malnutrition is common in active, severe A screening tool to detect IBD patients at risk of malnu-
IBD; however micronutrient deficiencies (vitamins, min- trition and nutrient deficiencies, rather than those who are
erals and trace elements) are seen even in patients with already malnourished, is therefore needed. Existing nutri-
mild disease or in clinical remission. Micronutrient defi- tion screening tools may be of limited use in the IBD out-
ciencies can lead to co-morbidities including anemia, patient population, as BMI and weight loss are often key
osteoporosis, thrombophilia, colorectal cancer, and poor
3
wound healing. Weight loss may not be the best measure
Corresponding Author: Natasha Haskey, Multidisciplinary
of nutrition risk in IBD, as emerging literature suggests
Inflammatory Bowel Disease Clinic, Division of Gastroenterol-
that IBD patients in remission have similar body mass
ogy, Department of Medicine, University of Saskatchewan,
4,5
indices (BMI) as healthy controls. As well, there is a
Royal University Hospital, 103 Hospital Drive, Saskatoon, Sas-
6-10
growing prevalence of obesity in the IBD population.
katchewan, Canada S7N 0W8.
A number of nutrition screening tools are available that
Tel: +001-306-290-7350; Fax: +001-306-844-1523
have been validated in a variety of populations including
Email: natasha.haskey@gmail.com; nhaskey@shaw.ca
11-13
medical, oncologic, and surgical patients. However,
Manuscript received 24 December 2016. Initial review complet-
routine nutrition screening is not commonly performed in ed 15 February 2017. Revision accepted 10 April 2017.
the IBD outpatient setting, resulting in under detection doi: 10.6133/apjcn.112017.01
Nutrition screening for inflammatory bowel disease 757
1,3-7 19
measures used in these tools. A recent systematic review oped from the literature, available tools, and clini-
indicates that BMI does not accurately predict body com- cian experience. The SaskIBD-NR Tool considers symp-
16
position in IBD and the growing prevalence of obesity toms (nausea, vomiting, and diarrhea), nutrient intake
6-10
in the IBD population, means that patients with nor- (food intake and food avoidance), and unintentional
mal or elevated BMIs may not be appropriately identified weight loss, all of which are all well-defined risk factors
as at risk of malnutrition with traditional screening tools. for malnutrition in the IBD outpatient population. Ques-
12
The malnutrition universal screening tool (MUST) tions pertaining to symptoms gauge IBD disease activity
has been validated in varied populations, including medi- (active or remission). Questions pertaining to nutrient
17
cal, surgical, and general gastroenterology patients. The intake screen for potential micronutrient deficiency.
tool focuses on BMI, weight loss, and the acute disease Questions relating to weight loss screen for potential pro-
effect. Although the MUST has not been validated specif- tein-energy malnutrition. Once a final version of the
ically in the IBD patient population, it has been reported questionnaire was defined, content validity of the tool
to be quick and easy to use which is desirable in demand- was evaluated by the Saskatchewan multidisciplinary
17,18
ing IBD outpatient settings. A recent publication IBD team which acted as a committee of experts. Subse-
showed that patient-administered MUST was comparable quently, the SaskIBD-NR Tool was piloted by dietetic
to healthcare practitioner-administered MUST, in the interns with five IBD patients to determine if the ques-
18
outpatient IBD clinic setting. However, potential limita- tions were clear and easy to understand.
tions of MUST are the emphasis it places on BMI, and
that it does not take into consideration recent nutrient Sample
intake and food avoidance, which are risk factors for mi- The study was conducted in the outpatient department at
cronutrient deficiencies. Royal University Hospital in Saskatoon, Saskatchewan,
A screening tool to detect nutrition risk that is quick Canada in the Multidisciplinary Inflammatory Bowel
and easy to administer and that identifies key nutrition Disease Clinic over a three-month period. A convenience
risk factors in the IBD outpatient population does not sample of 110 outpatients with IBD participated in the
currently exist in clinical practice. Given that elimination study. All participants in the study were ≥18 years, and
diets, food exclusion and micronutrient deficiencies are had an established diagnosis of IBD based on standard
common in IBD patients, and that BMI may not be an clinical, radiologic, endoscopic and histologic criteria.
accurate predictor of nutrition risk in this population, the Pregnant women with IBD were excluded.
aim of this study was to develop a reliable and valid nu- Prior to entering the treatment room, weight (kilo-
trition screening tool that would identify patients with grams) and height (meters) were completed by support
IBD at risk for malnutrition and potential nutrient defi- staff, and BMI was calculated (kilograms divided by me-
ciencies in the outpatient setting. ters squared). Patients were asked the questions on the
SaskIBD-NR Tool (Table 1) and the MUST (Table 2) by
MATERIALS AND METHODS the gastroenterologist, dietitian, or nurse practitioner as
Development of a nutrition screening tool for patients part of the patient’s regularly scheduled appointment.
with IBD Responses to each of the nutrition screening questions in
The Saskatchewan Inflammatory Bowel Disease – Nutri- the SaskIBD-NR Tool and MUST were given a score
tion Risk Tool (SaskIBD-NR Tool) is a locally developed (low, medium or high-risk categories). For the purposes
screening tool that was initially developed by three die- of analysis, patients falling into the ‘at risk’/‘medium
tetic interns. Key criteria for development of the nutrition risk’ and ‘malnourished’/‘high risk’ groups (≥1 for
screening tool were that it 1) be simple, quick and easily MUST and ≥3 for the SaskIBD-NR Tool) were combined
completed by all team members; 2) use data that was into one ‘pooled-risk’ group for each method of screen-
routinely available; 3) be non-invasive and economical; 4) ing. This method of combining risk groups has been pre-
17,19-21
could be incorporated into routine assessment; and 5) be viously used in similar studies. The prevalence of
valid and reliable. The dietetic interns met with the mul- nutrition risk using both screening tools was compared.
tidisciplinary IBD team (Registered Dietitian, two Gas-
troenterologists, Nurse Practitioner, Nurse Clinician, and Reliability of the SaskIBD-NR and the MUST screening
IBD Research Coordinator) to determine nutrition risk tools
factors that should be incorporated into a nutrition The Registered Dietitian and Gastroenterologist (RD/GI)
screening tool. assessment was chosen as the “gold standard” for deter-
The questions in the SaskIBD-NR Tool were devel- mining the actual risk of malnutrition. A major challenge
Table 1. SaskIBD-NR Tool
Nutrition screening item Score
1. Have you experienced nausea, vomiting, diarrhea or poor appetite “no symptoms”=0, “1-2 symptoms”=1, “≥3 symptoms”=2
for greater than two weeks?
2. Have you lost weight in the last month without trying? “no”=0, “unsure”=1, “yes”=see below
IF YES, how much weight have you lost? “<5 lbs”=0, “5-10 lbs”=1, “10-15 lbs”=2, “>15 lbs”=3
3. Have you been eating poorly because of a decreased appetite? “no”=0, “yes”=2
4. Have you been restricting any foods or food groups? “no”=0, “yes”=2
Total score: 0-2=low risk, 3-4=medium risk, ≥5=high risk.
758 N Haskey, JN Peña-Sánchez, JL Jones and SA Fowler
Table 2. MUST
Nutrition screening item Score
Step 1: BMI score “>20”= 0, “18.5-20”= 1, “<18.5”= 2
2
BMI=kg/m
Step 2: Weight loss score “<5”= 0, “5-10”= 1, “>10”= 2
Unplanned weight loss in past 3-6 months (% Score)
Step 3: Acute disease effect score “no”= 0, “yes”=2
Patient is acutely ill and there has been or is likely to be no
nutritional intake for >5 days
Total score: 0=low risk, 1=medium risk, ≥2=high risk
Table 3. Criteria used to determine nutrition risk: RD/GI assessment
Diagnosis (Crohn’s disease or ulcerative colitis)
Body mass index (BMI)
Unintentional weight loss
Presence or absence of symptoms (stools, vomiting, nausea, pain)
Location of disease, disease severity, concurrent conditions
Surgical history
Medications
Laboratory parameters (albumin, vitamin D, iron status, vitamin B12)
† †
Simple Colitis Activity Index (SCAI) , Harvey Bradshaw score (HBS)
Intake (appetite, food restriction)
†
When available.
to validating nutrition risk screening tools is the absence MUST and SaskIBD-NR screening tools. Using the
of single “gold standard” for identifying patients at risk pooled-risk groups, sensitivity, specificity, positive pre-
22
of malnutrition. The SGA is widely considered the gold dicted value (PPV), negative predictive value (NPV), and
standard in many studies pertaining to malnutrition. ROC area were computed, with their respective 95% con-
However, the SGA identifies patients who are already fidence intervals (95% CI), for the SaskIBD-NR and
moderately or severely malnourished, rather than those MUST screening tools. Inter-rater reliability was not
who are at risk of becoming malnourished. SGA exam- evaluated given that each evaluation was completed by a
ines overall nutrition intake versus specific food avoid- gastroenterologist, dietitian, or nurse practitioner. Statis-
2,23,24
ance which is common in the IBD population. As tical analyses were performed using IBM SPSS Statistics
well, a study assessing different indicators of malnutri- version 23 (SPSS Inc. Chicago, IL) and the diagti com-
tion in IBD patients found that although 74% of patients mand in STATA version 13 (Stata Corporation, College
were well nourished according to the SGA, these patients Station, TX).
had decreases in body cell mass and handgrip strength The University of Saskatchewan Research Ethics pro-
5
compared to controls. We therefore instead used a com- vided an exemption for ethics board review prior to ini-
prehensive assessment by the RD/GI as the “gold stand- tiation of the study. Patients provided informed consent
ard” for determining the patients who are at risk of mal- for the use of their data in the evaluation of these screen-
nutrition. ing tools. If patients were deemed at nutrition risk, by any
For each patient, the RD and GI completed a retrospec- method, they were referred to the RD for further assess-
tive chart review using the criteria outlined in Table 3 to ment.
determine nutrition risk. No formal scores were assigned
for each criterion. Rather, all the factors were taken into RESULTS
account by the RD and GI to determine if patients were The SaskIBD-NR Tool
either ‘not at risk’ or ‘at risk’ of malnutrition. The RD/GI The SaskIBD-NR Tool evaluates four components: gas-
assessment was completed within one week of the nutri- trointestinal symptoms, weight loss, anorexia, and food
tion screening. The RD and GI were not aware of the restrictions (Table 1). This nutrition screening tool was
scores of the SaskIBD-NR or the MUST until the chart specifically developed for patients with IBD. The com-
review (RD/GI assessment) was completed. Upon com- mittee of experts verified the content validity of the final
pletion of the chart review the patient’s scores on the version of the SaskIBD-NR Tool and approved it. In the
SaskIBD-NR Tool and the MUST were compared to the pilot, patients with IBD confirmed that the questions of
RD/GI assessment. To assess concurrent validity, the the screening tool were clear and understandable.
SaskIBD-NR Tool and the MUST score were compared.
Cohen’s kappa statistic was computed to measure Sample group
agreement between the SaskIBD-NR tool and RD/GI Demographics and clinical characteristics of IBD patients
assessment, as well as between MUST and RD/GI as- are summarized in Table 4. Mean age was 39 years
sessment. Receiver operating characteristic (ROC) (SD=15), 63 (57.3%) participants were female, 75
curves were also drawn using the actual scores of the (68.2%) patients had CD, and 35 (31.8%) UC. Mean BMI
Nutrition screening for inflammatory bowel disease 759
2
Was 26.4 kg/m (SD=5.8).
Table 4. Demographics and clinical characteristics
(n=110)
Reliability of the SaskIBD-NR and the MUST screening
Mean±SD, range/n (%)
tools
Age, years 39±15, 17-79
All participants were screened with the SaskIBD-NR
Gender
Tool, MUST and had a RD/GI nutrition risk assessment.
Female 63 (57.3)
Differences were observed in the prevalence of IBD pa-
2
Body mass index (kg/m ) 26.4±5.8, 17.7-43.2
tients at nutritional risk using these 3 methods of assess- †
Diagnosis
ment (Figure 1). The RD/GI assessment identified 20.9%
CD 75 (68.2)
(95% CI 13.7-29.7%, n=23) of the patients at nutritional Upper gastrointestinal 2 (2.7)
Ileal 21 (28)
risk. The SaskIBD-NR Tool classified 19.1% (95% CI
Colonic 21 (28)
12.2-27.7%, n=21) of the patients at some nutritional risk:
Ileocolonic 32 (42.7)
9 (8.2%, 95% CI 3.8-15%) at high risk and 12 (10.9%,
Perianal 11 (14.7)
95%CI 5.8-18.3%) at medium risk. In contrast, the
UC 35 (31.8)
MUST considered that only 15.5% (95%CI 9.3-23.6%,
Proctitis 7 (20)
n=17) of the patients were at some nutritional risk: 5 Left-sided colitis 16 (45.7)
Extensive colitis 12 (34.3)
(4.5%, 95%CI 1.5-10.3%) at high risk and 12 (10.9%,
‡
HBI score 4.2±6.7, 0-35
95%CI 5.8-18.3%) at medium risk.
§
SCCAI score 1.3±2.2, 0-9
The results of the nutritional screening tools were
Medication type for IBD (%)
compared with the results of the RD/GI assessment (Ta-
None 16 (14.5)
ble 5). A high and significant agreement was identified
5-aminosalicylic acid 32 (29.1)
between the SaskIBD-NR Tool and RD/GI assessment Corticosteroids 5 (4.5)
Immunomodulator monotherapy 17 (15.5)
(kappa 0.83, p<0.001), with good levels of agreement
Anti-TNF monotherapy 14 (12.7)
among both patients with CD (kappa 0.82, p<0.001) and
Other biologics 3 (2.7)
UC (kappa 0.84, p<0.001). Conversely, a lack of agree-
Immunomodulator + biologic 13 (11.8)
ment was observed between the MUST and RD/GI as-
Other combined schemes 10 (9.1)
sessment (kappa 0.15, p=0.12). This disagreement was
†
Percentages total more than 100% because some patients have
similar among patients with CD (kappa 0.14, p=0.27) and
been counted in more than one category.
UC (kappa 0.16, p=0.31). There was no significant
‡
Harvey-Bradshaw Index (HBI), n=67.
agreement between the SaskIBD-NR Tool and MUST §
Simple Clinical Colitis Activity Index (SCCAI), n=33.
(kappa: 0.18, p=0.06). A larger ROC area was observed
for the SaskIBD-NR tool (97.7%, 95% CI 95.5-98.3%) in
Sensitivity and specificity of the SaskIBD-NR Tool
comparison to the ROC area of the MUST (56.4%, 95%
was tested at different cut-off values to determine varia-
CI 46.7-66%) (Figure 2a).Using the pooled-risk group,
tions of this screening tool. This evaluation identified that
the ROC area of the SaskIBD-NR Tool was 90.2%
the chosen cut-off in this study (i.e., classifying IBD pa-
(95%CI 82.1-98.2%) versus 56.7% (95% CI 46.9-66.5%)
tients with a score of ≥3 as at risk of malnutrition and
for the MUST (Figure 2b).
those with a score of <2 at low risk of malnutrition) had
Figure 1. Prevalence of IBD patients at nutritional risk according to the RD/GI nutrition risk assessment, SaskIBD-NR Tool, and MUST.
no reviews yet
Please Login to review.
